Skip to main navigation
Skip to search
Skip to main content
VTT's Research Information Portal Home
Search content at VTT's Research Information Portal
Home
Profiles
Research output
Projects
Datasets
Research units
Research Infrastructures
Activities
Prizes
Press/Media
Impacts
Too Tiny to See: Hazardous Obstacle Detection Dataset and Evaluation
Topi Miekkala
, Samuel Brucker
, Stefanie Walz
, Filippo Ghilotti
, Andrea Ramazzina
, Dominik Scheuble
,
Pasi Pyykönen
, Mario Bijelic
, Felix Heide
BA4801 Automated vehicles
Torc Robotics
Mercedes-Benz AG
Princeton University
Research output
:
Contribution to conference
›
Conference article
›
Scientific
›
peer-review
Overview
Fingerprint
Fingerprint
Dive into the research topics of 'Too Tiny to See: Hazardous Obstacle Detection Dataset and Evaluation'. Together they form a unique fingerprint.
Sort by
Weight
Alphabetically
Computer Science
Point Cloud
100%
Autonomous Driving
100%
Evaluation Approach
100%
Support Analysis
100%
Graph Neural Network
100%
Depth Estimation
100%
Fair Comparison
100%
Estimation Method
100%
Feature Extraction
100%
obstacle detection
100%
Structural Similarity
100%
Texture Pattern
100%
INIS
detection
100%
evaluation
100%
depth
100%
datasets
100%
resolution
60%
comparative evaluations
40%
sensors
40%
range
40%
prediction
40%
accuracy
40%
distance
40%
lidar
40%
performance
20%
levels
20%
surfaces
20%
data
20%
metrics
20%
images
20%
texture
20%
shape
20%
cameras
20%
clouds
20%
graphs
20%
incidents
20%
extraction
20%
time-of-flight method
20%
neural networks
20%
ground truth
20%
Keyphrases
Dataset Approach
50%
Modality Agnostic
50%